PART TILTING IN CAPILLARY PART TILTING IN CAPILLARY--BASED SELF BASED SELF--ASSEMBLY: ASSEMBLY:

advertisement
PART TILTING IN CAPILLARYCAPILLARY-BASED SELFSELF-ASSEMBLY:
MODELING AND CORRECTION METHODS
S. Abbasi1, A.X. Zhou2, R. Baskaran1, 3, and K.F. Böhringer1
1Dept.
of Electrical Engineering, 2Dept. of Bioengineering, University of Washington, WA, USA,
3Intel Corporation, Chandler, AZ, USA
Abstract
We present a model and experimental results on tilt angle of microparts
in capillarycapillary-driven self
self--assembly.
assembly. The assembly is carried out in an
aqueous environment, using a heat curable adhesive for part
part--substrate
lubrication and mechanical bonding
bonding.. Silicon parts and substrate have
matching hydrophobic binding sites, which drive the assembly by
surface energy minimization
minimization.. Force balance analysis of an assembled
part leads to a model describing the dependence of tilt angle on
assembly parameters such as adhesive volume and waterwater-adhesive
interfacial tension.
tension. The effect of adhesive volume on tilt angle is
investigated experimentally.
experimentally. Tilt correction of the assembled parts is
achieved by providing external energy to the system via vertical
vibration..
vibration
Self--Assembly Process
Self
Correction Methods
•
The adhesive used for assembly is hydrophobic and heat curable, and
is composed of 97 wt.
wt.% triethyleneglycol dimethacrylate as a
monomer, and 3 wt.
wt.% benzoyl peroxide as the thermal initiator
•
Th assembly
The
bl template
t
l t is
i immersed
i
d in
i a hydrophobic
h d
h bi adhesive
dh i
and
d
pulled out into the water, because of surface energy minimization
and hydrophobicity of adhesive and substrate binding sites, the
adhesive selectively covers the hydrophobic binding sites
•
Parts are introduced to the template, and due to surface energy
minimization, they attach to the adhesive
adhesive--coated binding sites
•
After assembly completion, the adhesive is cured by heating the
water to 70ºC
70ºC for 2 hours, and mechanical bonding is achieved
•
Random direction of part approach during assembly causes vertical
tilt
•
Surface Evolver simulations show that the energy minimum of the
system
t
i in
is
i flat
fl t state
t t
•
Tilt correction can be achieved by external agitation which provides
enough energy to the system to transfer to the global energy
minimum (flat state)
•
Vertical vibration is used as external agitation by positioning the
assembly container on the stage of a speaker, connected to a signal
generator with a 15Hz,
15Hz, 9V sinusoidal wave for 2 minutes
Design
•
Wafer-scale 3-dimensional microelectronics with high interconnect
Waferdensity needs high part lateral size to thickness ratio:
ratio: 5×5×0.1mm3
silicon parts
•
Rectangular binding site allows for asymmetry and thus more
flexibility in the circuitry design
•
The length to width ratio of rectangular binding sites needs to be
more than 1.3:1 to get correct assembly orientations:
orientations: 3.65×
65×4.9mm2
binding site
Modeling and Experiments
•
Force balance analysis is used to model part tilt angle as a function of
assembly parameters
Before vibration
Vα = WL2 sin(α 2) cos(α 2) + W ( R 1 β − L sin(α 2) R 1cos( β ))
2
Fc = ΔPA = γ AW (
Fg , y = mg cos(α )
1
+
R1
1
R2
−1
)A
⎛ tρ g cos(α ) 1 ⎞
R1= ⎜⎜
− ⎟⎟
R2 ⎠
⎝ γ AW
−1
β = sin ( L sin(α / 2) / R1 )
After vibration
Fabrication
Substrate
•
Binding sites are patterned using Cr/Au evaporation and liftlift-off on
silicon substrate
•
A selfself-assembled monolayer (SAM) of thiol molecules which selectively
attaches to the gold is used to make binding sites hydrophobic (water
contact angle
angle:: 110
110º)
º) by soaking the patterned substrate in 1 mM
dodecanethiol in ethanol overnight
Part
•
Binding sites are patterned using Cr/Au evaporation and liftlift-off on
silicon wafer:
wafer: lab atmosphere exposed gold has water contact angle of
70º
70
º which is hydrophobic enough for our purpose
•
A grinding process from the back side of the wafer thins the silicon
wafer to 100μm,
100μm, and then the wafer is diced into individual parts
•
Tilt angle is measured as a function of adhesive volume
volume.. The tilt angle
increases by adhesive volume increment, hence tilt can be minimized
by volume minimization
•
Volume minimization needs very good control on dip
dip--coating
parameters
•
A model of tilt angle as a function of assembly parameters is proposed
•
Model and experimental results comparison shows fine compatibility
•
The control and minimization of part tilt in capillarycapillary-based selfselfassembly is investigated
•
A tilt correction method is demonstrated using external agitation
Model
Experiment
Conclusion
References::
References
[1] X. Xiong, Y. Hanein, J. Fang, Y. Wang, W. Wang, D.T. Schwartz, and
K.F. Böhringer, J. Microelectromech.
Microelectromech. Syst.
Syst., Vol
Vol..12,
12, No.
No.2, 2003.
2003.
[2] U. Srinivasan, D. Liepmann, and R.T. Howe, J. Microelectromech.
Microelectromech.
Syst.., Vol
Syst
Vol..10
10,, No
No..1, 2001
2001..
12th
th International
[3] K.L. Scott, R.T. Howe, and C.J. Radke, The 12
Conference an Solid State Sensors, Actuators and Microsystems,
Boston, 2003.
2003.
Acknowledgement
•
This work was supported by research grants from Intel Corporation. A. Z.
was supported by UWEB REU program at the University of Washington.
The staff of the Microfabrication Laboratory at the Washington
Technology Center provided help in cleanroom microfabrication
processes.
Download